© 2020 Arizent. All rights reserved.

FINRA Says Know Your Algorithms, or Risk Liability

Register now

The new FINRA report on digital investment advice makes it clear: If an advisor or broker is using a risk survey that maps to a model portfolio—be it a managed account or robo advisor technology for advisors— they should assess whether the algorithm is consistent with the firm’s investment process. Otherwise, they’re taking on the risk that the solution is not suitable.

If you are using someone else’s code or algorithms to onboard your clients and invest their money—without having the ability to interject yourself into the situation—you’re assuming a high degree of liability since you cannot attest to the suitability of that investment. You’re taking at face value that the algorithms are accurate.

That said—if you can demonstrate that you truly understand the investment process that is happening under the hood of this technology and can adequately discuss the offerings provided by the solution while demonstrating due-diligence, then there should be no problem. Given the complex nature of these models, this would likely require an outside consultant or the advisory firm hiring an in-house technology expert, which may be expensive or onerous.


Technology solutions are meant to make lives easier as well as potentially reduce costs. The whole idea behind RIAs and small broker-dealers leveraging technology is to automate and simplify their lives so that they can focus on the end client. Documenting processes, portfolio performance and data aggregation all come without inherent risks given they don’t fall under the umbrella of digital advice. These won’t have associated costs.

Now, if smaller broker-dealers are happily using financial planning and investment technology offerings without doing their due-diligence, there will likely be an added cost of hiring a third-party inspector, an in-house consultant, or another unbiased assessment of the consistency, accuracy and suitability of these solutions.  

The reason that FINRA can say that they aren’t changing the existing broker-dealer regulation obligations is that there has been a focus on suitability. FINRA has gone after several firms in the last two years for not being able to demonstrate suitability. The rules are not different per se, but they are spelling them out now to ensure that no one is confused about the implications.


Current technology solutions do not ask enough questions or delve deep enough to determine what makes each individual client unique. Asking simple survey questions about gender, age, desired retirement age and current cash flow does not tell you anything about an individual’s human capital.

Two 35-year-olds with 30 years until retirement, who earn the same income, should not be assumed to have the same, or even similar, suitable investments. You need to ask questions about their health, sectors of excess exposure, and their career stability to get a better sense of what their investment profile should look like. If a live advisor is completely disintermediated by the technology solution they are employing, there is no room for the human element of making sure that there are no errors or conflicts in the data collection.

Simply going from questionnaire to investing without a second set of eyes is incredibly risky for both the client and the advisor. Specifically, the advisor is implicitly signing off on a process that they have no control or oversight on—which is an increasingly precarious position to find oneself in. If there is not enough specificity on the questionnaires to address what makes an individual unique, then the investment solution by its very nature cannot possibly claim to be a bespoke or tailored solution.

If there is no oversight to ensure that there were no conflicting responses on the customer profile, (interplay between risk they should take versus risk they want to take) then a client may find themselves in an investment that isn’t suitable for one reason or another. These are all things that need to be taken into consideration.


Ultimately, all advisors or registered reps have a responsibility to know what they are investing their clients’ money in, as well as to know the end client. This has always been true of compliant, successful, upstanding fiduciaries.  

If an automated front-end controls all, or part of the process from onboarding to investing with the advisor only supervising, then the advisor can’t really claim to be involved in know-your-customer or the security selection process. That is painting it in the broadest strokes and there is certainly some nuance there.  

Advisors can rely on tools that help them get to know the client better and serve as an engagement apparatus to develop the know-your-customer process. Tools that develop the conversation and demonstrate that you are living up to the highest standards of suitability should definitely be embraced as they don’t run into the pitfalls of an entirely automated workflow.

These recommendations of assessing digital tools hit at some important, if not obvious, points (as seen from FINRA below). If the tool is found lacking in any of these areas, then it is not properly considering the suitability of suggested investments. 

  • Does the tool seek to obtain all of the required investment profile factors? 
  • If not, has the firm established a reasonable basis to believe that the particular factor is not necessary? 
  • How does the tool handle conflicting responses to customer profile questions? 
  • What are the criteria, assumptions and limitations for determining that a security or investment strategy is suitable for a customer? 
  • Does the tool favor any particular securities and, if yes, what is the basis for such treatment? 
  • Does the tool consider concentration levels and, if so, at what levels (e.g., particular securities, class of securities, industry sector)?

Min Zhang is CEO and co-founder of Totum Wealth.

Read more:

For reprint and licensing requests for this article, click here.